Thesis_MES_2021_McNamaraC.pdf
Media
Part of COHO SALMON RESPONSE TO CHANGES IN STREAMFLOW AND HARVEST PRESSURE IN BIG BEEF CREEK, WA
- extracted text
-
COHO SALMON RESPONSE TO CHANGES IN
STREAMFLOW AND HARVEST PRESSURE IN
BIG BEEF CREEK, WA
by
Caitlin McNamara
A Thesis
Submitted in partial fulfillment
of the requirements for the degree
Master of Environmental Studies
The Evergreen State College
June 2021
©2021 by Caitlin McNamara. All rights reserved.
This Thesis for the Master of Environmental Studies Degree
by
Caitlin McNamara
has been approved for
The Evergreen State College
by
________________________
John Kirkpatrick, Ph. D. –
Member of the Faculty
________________________
Date
ABSTRACT
Coho salmon response to changes in streamflow and harvest pressure in Big Beef Creek,
WA
Caitlin McNamara
Several factors were considered to look at population dynamics of wild stock coho
salmon found in Big Beef Creek, Washington. Big Beef Creek is a rain dominated system
and home to long term monitoring of coho salmon through weir operations run by the
Washington Department of Fish and Wildlife. The date in which the mode of coho
salmon returning to the weir was found to be changing over the past 20 years to an earlier
date. Coho often face delays due to lack of streamflow and harvest pressure from the
terminal net fishery. Because of the combined effects coho salmon at Big Beef that pass
above the weir to spawn are smaller and may be of less fitness.
Table of Contents
TABLE OF CONTENTS ........................................................................................................................... IV
LIST OF FIGURES ..................................................................................................................................... V
LIST OF TABLES .................................................................................................................................... VII
ACKNOWLEDGEMENTS .................................................................................................................... VIII
CHAPTER ONE: INTRODUCTION ..........................................................................................................1
STUDY AREA ...............................................................................................................................................2
COHO SALMON PHENOLOGY ........................................................................................................................4
STREAMFLOW TIMING ..................................................................................................................................5
HARVEST .....................................................................................................................................................6
REGIONAL MARK PROCESSING CENTER AND CODED WIRE TAGGING PROGRAM ........................................7
CHAPTER TWO: LITERATURE REVIEW ............................................................................................9
HABITAT SELECTION IN COHO SALMON...................................................................................................... 10
STREAM FLOW TIMING ............................................................................................................................... 12
IMPLICATIONS ............................................................................................................................................ 15
ALTERNATIVE METHODS............................................................................................................................ 18
PHENOTYPIC PLASTICITY AND EVOLUTIONARY CHANGES: ......................................................................... 18
CHAPTER THREE METHODS: .............................................................................................................. 22
STREAMFLOW: ........................................................................................................................................... 22
WEIR ......................................................................................................................................................... 23
HARVEST ................................................................................................................................................... 24
PEAK ADULT MIGRATION AND LENGTHS: ................................................................................................... 26
JUVENILE OUTMIGRATION AND TAGGING ................................................................................................... 27
STATISTICAL ANALYSIS ............................................................................................................................. 28
CHAPTER FOUR RESULTS: ................................................................................................................... 30
SEASONAL FRACTIONAL FLOW FOR BIG BEEF: .......................................................................................... 33
HARVEST ................................................................................................................................................... 35
TIMING ...................................................................................................................................................... 38
SIZE ........................................................................................................................................................... 40
MULTIPLE REGRESSION ............................................................................................................................. 45
JUVENILE OUTMIGRATION TO ADULT MIGRATION ...................................................................................... 46
CHAPTER FIVE DISCUSSION................................................................................................................ 47
STREAMFLOW ............................................................................................................................................ 47
PACIFIC NORTHWEST HYDROCLIMATOLOGY ............................................................................................. 49
MIGRATION TIMING ................................................................................................................................... 49
MIGRATION TIMING RELATIONSHIPS .......................................................................................................... 51
SIZE ........................................................................................................................................................... 52
CHAPTER SIX CONCLUSION: .............................................................................................................. 53
BIBLIOGRAPHY ....................................................................................................................................... 56
iv
List of Figures
FIGURE 1: BIG BEEF CREEK HIGHLIGHTED IN GREEN WHICH SHOWS THE 18KM STRETCH
AND DRAINS INTO HOOD CANAL. LINE BREAK CORRESPONDS TO LAKE SYMINGTON.
3
FIGURE 2 MAP OF DRAINAGE AREA TO STREAMFLOW GAUGE LOCATED ON BIG BEEF CREEK.
_________________________________________________________________ 23
FIGURE 3 MARINE 12 FISHING AREA, HIGHLIGHTED IN ORANGE CROSS MARKING. MAP
COURTESY OF THE WASHINGTON DEPARTMENT OF FISH AND WILDLIFE. _________
25
FIGURE 4 STREAMFLOW FOR SEASONAL DURATION OF AUGUST TO DECEMBER FOR A. 2000,
B. 2010, AND C. 2020. WHERE RED IS THE SAME IN EACH, BEING THE TWENTY YEAR
AVERAGE AND BLUE INDICATES EACH YEARS FLOWS
________________________ 31
FIGURE 5 ADJUSTED PLOT TO COMPARE THE EFFECTS OF THE 2003 EARLY FLOODING EVENT
_________________________________________________________________ 33
FIGURE 6 SEASONAL FRACTIONAL FLOWS FOR A. 2000- 2020 AND B. HISTORICAL DATA
1971-1980 ________________________________________________________ 33
FIGURE 7 CENTRAL MASS FLOW TIMING FOR BIG BEEF CREEK OVER THE YEARS 2000-2020
_________________________________________________________________ 35
FIGURE 8 TOTAL ADULT COHO RETURN TO BIG BEEF CREEK THAT ARE PASSED UPSTREAM
TO SPAWN OVER THE PAST 20 YEARS
____________________________________ 38
FIGURE 9 PEAK RETURN OF COHO AT BIG BEEF CREEK THAT ARE PASSED UPSTREAM TO
SPAWN. WHERE DATE IS IN JULIAN OR CALENDAR DAY.
______________________ 40
FIGURE 10 BOX PLOT OF LENGTHS OF COHO THAT ARE CAUGHT IN THE TERMINAL NET
FISHERY AND THOSE THAT ARE PASSED UPSTREAM. BOX PLOTS FURTHER BROWKN
DOWN INTO SEX, WHERE HARVEST REPRESENTS THOSE THAT ARE CAUGHT IN THE
v
TERMINAL NET FISHERY AND FEMALE AND MALE ARE THOSE THAT WERE PASSED
UPSTREAM. LASTLY, TOTALS REPRESENT BOTH MALE AND FEMALES COMBINED THAT
WERE PASSED UPSTREAM. _____________________________________________
40
FIGURE 11 FEMALE COHO LENGTH TREND OVER TIME, P VALUE OF 0.02725. __________ 42
FIGURE 12 MALE COHO LENGTH TREND OVER TIME, P VALUE 0.005295. _____________ 43
FIGURE 13 TOTAL LENGTH OF COHO SALMON THAT ARE PASSED UPSTREAM AND ARRIVE TO
SPAWNING GROUNDS ON BIG BEEF CREEK. P VALUE 0.008715
________________ 44
FIGURE 14 LENGTHS OF COHO CAUGHT IN THE TERMINAL NET FISHERY, P VALUE 0.04103.
_________________________________________________________________ 45
vi
List of Tables
TABLE 1 LIST OF PACIFIC SALMONIDS
FOUND IN BIG BEEF CREEK ..................................... 4
TABLE 2 PERCENT OF COHO HARVESTED IN TERMINAL NET FISHERY FOR EACH YEAR ....... 36
TABLE 3 SPEARMANS RANK CORRELATION FOR UPSTREAM FISH PASSAGE AND HARVEST TO
STREAMFLOW DISCHARGE RELATIONSHIPS. ................................................................ 37
TABLE 4: STATISTICAL SIGNIFICANCE RESULTS (P VALUES) OF MULTIPLE REGRESSION. .... 46
vii
Acknowledgements
I am grateful to have been able to work on this project and have the support of my
supervisor, Clayton Kinsel, fish biologist for WDFW. Thank you for being available to
me, answering questions, and discussing all my ideas. I am also very grateful to my thesis
advisor, John Kirkpatrick, thank you for all your input and help, not only with my thesis,
but throughout my graduate education at Evergreen. Richard Iverson, for supporting me
throughout this process, allowing me time and space to make sure I was successful in the
program, while taking care of our dogs and providing study meals. Lastly, I thank my
friends, family, and peers for all their support and encouragement throughout my
graduate education.
viii
Chapter One: Introduction
Salmon in the Pacific Northwest are an ecologically and economically important
species as they support significant commercial and artisanal fisheries and are sensitive
environmental indicators (Drenner et al., 2012). Factors that threaten the success of the
species are of major concern as Pacific salmon also provide commercial and tribal
cultural value (Ogston et al., 2015). There are seven species of Pacific Salmon
(Oncorhynchus spp.) that spawn in rivers on the west coast of North America and all of
these populations have declined significantly resulting in increased research and
monitoring efforts (Ford et al., 2008). Causes for population decline are frequently
attributed to habitat loss and overharvest. Artificial propagation of salmon is conducted
with the intent to supplement populations. There are several negative effects of this
artificial propagation that have been thoroughly researched. Hatchery practices can result
in domestication, and evolutionary models suggest that this results in the decline of
fitness within the natural populations they are supplementing (Ford et al., 2008). Wild
females and males have been found to have significantly greater sexual characteristics
than hatchery fish, have higher reproductive success, and higher smolt survival even in
poor marine conditions compared to hatchery fish (Beamish et al., 2012; Crozier et al.,
2008; Fleming & Gross, 2011). Therefore most management programs prioritize the
protection and conservation of wild salmon specifically (Heard, 2012; Irvine, 2009).
Criteria for assessing and identifying salmonid stocks as well as managing them
includes understanding location, timing, and abundance. Selecting the best available
habitat for reproduction, arriving to spawning grounds early, and body size are all
1
specific strategies and features that determine success for salmonids. Salmon also often
respond strongly to environmental disturbances which can cause evolutionary changes for
the fish within relatively few generations. For the purposes of this study, we will focus on
wild stock coho salmon (Oncorhynchus kisutch) found in Big Beef Creek, Washington
and population dynamics in response to synchronous environmental and anthropogenic
pressures.
Study Area
Big Beef Creek drains into Hood Canal at (47°39’N, 122°46’W) (Figure 1.). The
region is characterized by a depressed, glaciated area which is now partially submerged
(Kennedy et al., 1981). Ethnographic research has indicated the area to be within
traditional Twana (Skokomish Indian Tribe) territory (Kennedy et al., 1981). Between
1857 and 1886, during the Westward Expansion, the area was settled and used for
logging and dairy farming (Kennedy et al., 1981). This large-scale anthropogenic activity
resulted in major disturbances to the natural environment in Big Beef. It is also due to the
presence of cattle during this time of settlement that Big Beef Creek earned its name.
Big Beef Creek is used as an indicator stream for long-term ecological studies and
has been monitored by the Washington Department of Fish and Wildlife for over 30 years
(Kodama et al., 2012). A weir is located at the mouth of the stream which has allowed
researchers to monitor migrating fish passage. Pacific salmon and trout species that are
found migrating in and out of Big Beef Creek are steelhead, cutthroat, coho, and chum
(Table1.) Only wild stock coho salmon that return to the weir are passed above upstream
to spawn. Hatchery-origin Coho Salmon are identified by mark status (adipose clip) and
2
are not allowed above the weir (Kinsel & Zimmerman, 2011). This ensures the genetic
integrity of the wild stock coho salmon.
Figure 1: Big Beef Creek highlighted in green which shows the 18km stretch and drains
into Hood Canal. Line break corresponds to Lake Symington.
3
Table 1 List of Pacific Salmonids found in Big Beef Creek
Common name
Scientific name
Steelhead
Oncorhynchus mykiss
Cutthroat
Oncorhynchus clarkii
Coho
Oncorhynchus kisutch
Chum
Oncorhynchus keta
Coho salmon phenology
Coho salmon have been historically abundant in Washington state and typically
select smaller streams and tributaries to reproduce in (Groot & Morgolis 1991). They
spend 18 months of their initial life cycle in fresh water and 18 months in the ocean
where rapid growth occurs. The initial life cycle is crucial and this length and
development depends on temperature unit and seasonal accumulation of heat energy
(Dittmer, 2013).
The phenology of coho salmon is highly dependent on long term-averages in
abiotic conditions such as precipitation and flow found (Crozier et al., 2008). These flows
dictate accessibility to spawning grounds for adults who return to spawn in native
streams. The reproductive migration for coho salmon begins in the fall. At Big Beef
4
Creek, the weir is typically actively fishing and collecting data on fish from the end of
August to the end of December annually, depending on weather conditions. According to
the 2009 Washington Department of Fish and Wildlife (WDFW) report, the first coho
returning to Big Beef Creek were detected in mid-September and the last returning
individual returning was detected on November 25th (Kinsel & Zimmerman, 2011).
Similar to reports by other researchers, a majority of these arrived after the first few rains
that were significant enough to impact flow level of the creek.
Date is a crucial aspect of the success of the species since females determine the
offspring’s environment exclusively by spawning site and date (Anderson et al., 2010).
Change in the arrival time to spawning ground or peak migration are phenological
markers for salmon and deviations over time can be indicators of changing environmental
conditions or anthropogenic stressors.
Streamflow timing
Warming climate can change the hydrologic cycle across temporal and spatial
scales (Kam et al., 2018). Big Beef Creek is within a rain dominated watershed where
flows are dictated by precipitation and significant delayed streamflow in recent years has
been anecdotally noted by researchers in Big Beef Creek, alongside delayed coho salmon
returns. Salmon have physiological responses to a multitude of factors within the
freshwater and marine environment. For example, temperature also plays a strong role in
determining migration as well as poor marine conditions.
This is crucial to consider when addressing the management for salmon.
Phenological traits are generally heritable in salmonid populations and it has been
5
previously hypothesized that microevolutionary changes in migration timing may be one
mechanism salmon populations use when faced with climate change (Kovach et al.,
2012). Coho salmon have a 2-3 year life cycle which makes them ideal candidates when
examining evolutionary changes and shift in migration timing potentially related to
changing environmental conditions or climate change.
Harvest
Simultaneous adult returns, commercial salmon fisheries are conducted in Hood
canal. One fishery that specifically targets coho and as bycatch in the commercial chum
fishery. This provides the fishery two different opportunities to catch coho as they may
become bycatch in the chum fishery. The timing of this fishery occurring seasonally and
annually has the ability to truncate the run of coho returning to Big Beef Creek.
Differential mortality patterns from harvest of wild populations can result in a decrease in
density, mean ages, and mean lengths of individuals (Kendall & Quinn, 2017).
Furthermore, this harvesting gear that is used often selectively removes larger individuals
and this results in a disruption of age and length at maturation among those that do
survive to reproduce (Kendall & Quinn, 2017). Reports have shown, 30-50% of
exploitation rates are seen within this fishery and fluctuating even to 60- 90% (Kinsel &
Zimmerman, 2011; Russell et al., 2018). The fisheries that are not targeting coho salmon
may also present a threat as species that are bycatch and released still suffer post-release
mortality rates that can be substantial (Gale et al., 2011; Raby et al., 2018). Exploitation
rates of coho salmon in Big Beef Creek can be precisely estimated because all fish are
coded-wire tagged at the weir upon outmigration.
6
Regional Mark Processing Center and Coded Wire Tagging program
The coded wire tag (CWT) was introduced to Alaska, British Columbia,
Washington, Idaho, Oregon, and California in the late 1960’s as an alternative to fin clip
and external tag for identification of anadromous salmonids (Nandor et al., 2010). CWTs
are small ~ 1 mm long and contains either a numeric code or binary code that is unique to
a specific region and it is widely used by federal, state, and tribal fishery managers. This
tag is inserted into juvenile salmonids and sits in the nasal cartilage of the fish. The use of
CWT remains the most important tool for salmonid research and are most frequently used
in studies examining multiple life stages as they allow management to gain insight on
ocean distribution patterns, fishery impacts, and survival rates for Pacific salmon
(Drenner et al., 2012; Nandor et al., 2010).
The Regional Mark Processing Center (RMPC) is designated by law to house and
maintain the CWT database in the U.S. and to be the designated site for sharing data with
Canada (Nandor et al., 2010). More specifically the RMPC manages data by (1)
maintaining and upgrading regional database for all CWT releases and recoveries, plus
release data for fish groups given other types of marks, (2) ensuring that reported data
meet established format standards and pass validation procedures, (3) developing and
maintaining on-line computer applications for querying and reporting from the database,
(4) providing electronic copies of data sets upon request, and (5) implementing
recommended changes in the regional database exchange formats to meet expanding
requirements for new information (Nandor et al., 2010).
7
Researchers have often used this data from coded wire tagging which is provided
by the RMPC. Specifically, this data is used to estimate ocean distribution patterns and
marine survival. Studies that have worked with this data use coastal marine fisheries as
samplers of CWTed coho salmon to investigate ocean distribution patterns (Weitkamp &
Neely, 2002). Furthermore, these recoveries were used to determine the movements of
adult coho salmon in coastal areas as they returned to their natal streams. This method is
also routinely used at Big Beef Creek to determine movement and marine survival. This
is an efficient method as sampling effort is broad enough to support statistically
significant findings. Furthermore, fisheries are generally targeting salmon when and
where they are aggregated and therefore easily caught (Weitkamp & Neely, 2002). Coho
salmon are also considered a less complex salmon species to estimate marine survival as
their life history is relatively consistent (Cochran et al., 2019). Most fish returning to
spawn each year are from the same cohort of out-migrating smolts (Cochran et al., 2019)
therefore, appropriate forecasts for each year may be made. It is this data and forecasts
that fisheries managers use when attending the North of Falcon Meeting and assessing
the SaSI, as mentioned previously.
8
Chapter Two: Literature Review
Big Beef Creek is part of the Hood Canal stream complex. The entire 18 km of
Big Beef Creek habitat consists of 8km upstream from Lake Symington and 10km
downstream to the mouth (Quinn & Phil Peterson, 1996). Lake Symington is a man-made
lake that was constructed in 1970 by installing a 10 meter dam which includes a fish
ladder for passage (Quinn & Phil Peterson, 1996). There has been extensive research and
restoration efforts in attempts to create a more dynamic and beneficial habitat for juvenile
coho and their 18-month initial freshwater stage at Big Beef Creek. This research is part
of the Intensively Monitored Watersheds (IMW) program where Big Beef is used as a
treatment stream to determine the effects of habitat restoration to fish production. Large
woody debris was added to Big Beef Creek in order to create a more dynamic stream
network and provide valuable habitat to salmonids at all life stages, with a focus on the
benefits to coho salmon. A dike was also removed which opened up the lower floodplain
for salmon use. The objectives for the IMW project are to (1) estimate abundance of coho
parr and parr-to-smolt survival in all four creeks, (2) estimate juvenile production of
coho, (3) compare timing of juvenile outmigration among watersheds, (4) determined
escapement of coho and chum into Big Beef Creek, (5) describe spawning distribution
9
and timing of coho in all four creeks, and (6) estimate harvest rate and marine survival of
Big Beef Creek coho (Kinsel & Zimmerman, 2011). This serves as an excellent example
of the effort and focus on freshwater studies for coho salmon and management. The
freshwater lifecycle is frequently studied as there are inherent difficulties with studying
salmon in the marine environment (Drenner et al., 2012). This leaves the marine phase of
the lifecycle somewhat limited despite being recognized as a critical stage and resulting
in a decline to salmon populations (Drenner et al., 2012).
There are gaps in research on the adult coho that return to Big Beef and shifts in
migration timing and body size. There are studies available to the shifts in migration
timing and evolution within salmonids. Often these shifts are linked to pressure from
terminal net fisheries and where salmon are typically size selected, resulting in negative
size trends. Salmon also are expected to face changes due to climate change; some
populations may already be experiencing this. Research on these effects and changes to
salmon and coho salmon specifically will be expanded on further in this literature review.
Habitat selection in coho salmon
Habitat selection is critical as it represents a behavioral adaptation that is assumed
to increase individual fitness (Clark et al., 2014). Habitat selection within the stream for
redd construction as well as the surrounding environment is critical as salmon exhibit
limited parental care (Clark et al., 2014). It is possible that a poor selection in
reproduction site can result in a complete loss of the females contribution to the next
generation (Clark et al., 2014). Coho salmon are typically found in smaller creeks, rivers,
and tributaries. Structurally complex systems with large woody debris, root wads,
10
vegetation, and gravel bed are ideal areas that support coho reproduction. Researchers
have found that about 85% of redds built by coho salmon occur in areas where the
substrate contains gravel size of 15cm in diameter (Groot & Morgolis 1991). Furthermore
redd construction has been typically consistent with maximum stream discharge periods
(Clark et al., 2014). Three variables that were highlighted as significant factors by
researchers studying habitat selection by female coho were (1) distance to nearest pool,
(2) depth at the tailspill, and (3) maximum stream depth (Clark et al., 2014). These are all
factors that will potentially affect success of offspring and provide protection for both the
spawning adults and juveniles that emerge. Stream depth is also especially important as
selecting these areas may further protect fish from predation.
Habitat selection is also a behavioral adaptation that can occur in females further
contributing to evolutionary changes seen in coho (Clark et al., 2014). This adaptation is
often overlooked or not mentioned in other studies that analyze evolutionary adaptations
of coho salmon. Researchers have frequently studied the demographics of habitat loss
and habitat loss specifically linked to anthropogenic processes however the evolutionary
consequences rarely studied (McClure et al., 2008). A large reduction in habitat can, (1)
reduce a systems capacity which could result in a reduction in effective population size
and (2) decrease genetic variability within the system (McClure et al., 2008). All
anadromous fish have experienced a dramatic change in habitat sites and accessibility
which has resulted in extirpation of entire runs or evolutionary changes that result in
dramatically altered selective regime (McClure et al., 2008). It is the loss of habitat is
largest threat to endangered species in the United States (McClure et al., 2008).
Researchers and literature have shown us the importance of habitat selection and we
11
know that extensive habitat restoration has been done on Big Beef Creek by introducing
large woody debris to create a more complex stream system which benefits coho even in
its juvenile lifestage. There are however gaps in research to the evolutionary changes in
habitat selection from coho salmon.
Stream flow timing
The biological processes and life stages of salmonids are driven by phenological
traits. This is seen when river temperatures rise in the spring, signaling fry to emerge, and
again in the fall when river flows and precipitation create access and cool oxygenated
water. River entry is dominated by temperatures and associated flow and discharge rates.
More specifically, return migration to the freshwater environment appear to be in two
phases: an initial phase with navigation from feeding areas towards the coast and
secondly more precise orientation in coastal waters (Davidsen et al., 2013). River entry
has also been inherently linked to larger and older salmonids entering prior to younger
and smaller salmon (Harvey et al., 2017). Researchers have also noted trends where
females enter the river system prior to males (Harvey et al., 2017; Kodama et al., 2012).
Streamflow timing may be highly affected by climate change. Several studies predict
higher global averages, resulting in a warmer atmosphere which promotes grater
hydrologic extremes, more severe drought in the summer and more intense precipitation
and flooding in the winter (Crozier et al., 2008). There have also been extensive studies
into the earlier streamflow timing associated with earlier snowmelt (W. D. Burke &
Ficklin, 2017; Ficklin et al., 2013; Stewart et al., 2005). However, gaps in literature
frequently occur when looking at rainfall dominated basins in coastal regions. An
increase in temperature is also expected to having a dramatic effect on streamflow as well
12
as precipitation. Studies in the Colorado River Basin suggest that water availability will
significantly decrease with streamflow reductions of 30% (Ficklin et al., 2013)
There is an extensive body of evidence that the river entry by salmonids is linked to the
amount of flow and discharge that is coming from the natal stream. There is also
increasing research conducted on the change in timing among rivers and streams from
rain dominated systems to snowmelt systems (Kam et al., 2018; N. Mantua et al., 2010;
Stewart et al., 2005). These changes in timing will undoubtably result in changes to
migration timing of salmon. Researchers have also found that changes in streamflow
timing and the time in which salmon enter the river appear are more important in smaller
stream systems (Davidsen et al., 2013). An example of smaller stream systems is known
as intermittent streams. Intermittent streams only flow for a portion of the year and make
up for 65% of streams found in the western U.S. (Wigington et al., 2006) and coho
salmon spawning habitat is frequently found in intermittent stream systems. A study
conducted in the West Fork Smith River in Oregon, U.S. showed that 21% of coho
salmon spawned in neighboring intermittent streams (Wigington et al., 2006). This
coincides with other researchers findings of coho persisting at higher rates in intermittent
streams than mainstem perennial reaches (Larsen & Woelfle-Erskine, 2018). There are
several small intermittent tributaries which provide ideal habitat for coho salmon both
above and below the Lake Symington on Big Beef Creek. The researchers that survey
Big Beef Creek specifically have noted that given the chance coho will largely be found
in these intermittent tributaries, however they frequently do not have enough water flow
to give them access.
13
Large scale climate studies have analyzed trends in ocean conditions as well as
streamflow records across the Pacific North West and note that understanding climate
change implications to fisheries management is critical because salmon production goals
may simply not be attainable when environmental conditions are unfavorable (Mantua et
al., 1997). These reductions in flow however will certainly reduce the availability of
spawning habitat for salmon populations that spawn early in the fall (Mantua et al.,
2010). This reduction of spawning habitat will be further emphasized by the elimination
of intermittent tributaries which may remain dry all season, limiting access and site
selection by coho.
A study conducted by Mantua et al., 2010 classifies Washington’s watersheds into
snowmelt dominant, transient, or rainfall dominant based. The Hood Canal stream
network falls into the rainfall dominated category (Figure 1) (Mantua et al., 2010). These
watersheds are predicted to face large changes in the coming decades and a trend in
delayed stream flow timing has already been observed. Most of Washington's river basins
are projected to experience reduced streamflow in summer and early fall that results in
extended period of low summer flows, while rainfall-dominant basins are projected to
have substantially lower base flows ( Mantua et al., 2010).
Fish that arrive early in the season frequently experience inaccessibility to
smaller tributaries that provide optimal spawning habitat. Researchers have found that the
flow in Big Beef Creek are dominant from rains that occur between November and
March (Quinn & Phil Peterson, 1996). More recently these averages have been published
as being from October to mid-November (Kodama et al., 2012). This fluctuation is
precipitation and streamflow is a driving factor for this study in looking at long term
14
trends. Precipitation and stream flow may result in fluctuations in selection as well as
affect timing of return fish (Kodama et al., 2012). This was seen by in a study where 80%
of the 2006 cohort in Big Beef Creek was delayed until the beginning of November
(Kodama et al., 2012). In the 2007 return migration of this study, adults were found to
return in peaks ranging from the beginning of October to the end of November.
Furthermore, the male to female ratio was higher in 2007 suggesting a directional
selection for larger 3-year-old males. For females in this study, a larger body size and
directional selection was favored in 2006 and intermediate size in 2007. This was
determined based on reproductive success. This research along with that of others suggest
that there is fluctuations in mode, directional selection and strength of selection on return
date in both sexes of salmon (Anderson et al., 2010; Kodama et al., 2012).
The salmon fishery in the Pacific Northwest is generally managed by
implementing gear restrictions and time and area closures, which limit the amount of
fishing opportunity available (Vander Haegen et al., 2004).
Implications
The salmon fishery in the Pacific Northwest is generally managed by
implementing gear restrictions and time and area closures, which limit the amount of
fishing opportunity available (Vander Haegen et al., 2004).
The strain of commercial harvest on salmon has been well documented. There are
however salmon fisheries that have been noted as archetypes of sustainable resources and
management, such as the Bristol Bay, Alaska sockeye fishery (Atlas et al., 2021). Salmon
are harvested during their oceanic feeding migration and numerous different populations
15
of salmon are caught at once. This is seen in Big Beef Creek coho as well, although many
are caught near the mouth or in the marine area nearby. The commercial fishery here
typically uses gillnet which are known for being size selective, but also can result in
delayed mortality from fish that experience unobserved entanglement or are released for
conservation concerns (Baker et al., 2011; Bass et al., 2018). Beach seine are arguably
the better method as post capture release survival rates for salmon are 95% compared to
40% in the gill net sets (Bass et al., 2018). However, this is only true when proper
handling techniques are used. Fishermen may frequently allow fish to stay in nets that are
pulled onshore and wait until fish become less active so they are easier to handle.
A study conducted by Vander Haegen et al., 2004 studied Chinook salmon
caught in two different size nets, an 8 inch gill net and a 5.5 inch gill net. Nearly every
fish from the study retained net marks or damage on the body with the 5.5 inch net
producing all marks around the snout and the 8 inch resulting in net marks and damage
on the body. This body damage was seen to be severe and resulting in a large loss of
scales, slime, or scars and only 57% post capture successfully recovered (Vander Haegen
et al., 2004). This is similar to post capture release survival found in a study conducted on
sockeye salmon in Bristol Bay by Baker et al. 2011, however these researchers
emphasize that this estimate is conservative and post release mortality may be up to 74%.
Differential mortality patterns from harvest of wild populations can have
significant ecological effects, including reductions in density with associated increases in
growth and decreases in mean ages and lengths of individuals (Kendall & Quinn, 2017).
Furthermore, harvesting gear often selectively removes individuals with respect to length
(Kendall & Quinn, 2017). Fishery selection may lead to genetic changes in life-history
16
traits which may be harder to reverse than changes associated only with phenotypic
plasticity (Law, 2000). There is a body of evidence of this effect for the last century on
selective harvest resulting in a shift towards smaller sized fish and even a decreased age
at maturity. Additional issues with size selective harvest include:
(1) decreased fecundity
(2) increased sexual dimorphism
(3) lowered reproductive rates
(4) reduced yield
(5) increased variability in abundance
(6) stock collapse
283 years of size selection patterns were quantified from nine Alaskan sockeye salmon
fisheries and direction and size selection patterns were analyzed (Kendall & Quinn,
2017). In 72% and 84% of the years the fisheries caught larger than average male and
female fish, respectively, leaving smaller fish to spawn (Kendall & Quinn, 2017). The
results of this study also showed that nonlinear selection differential values for males
were significantly larger than females, indicating males experienced more disruption than
females (Kendall & Quinn, 2017).
In a study conducted by Ohlberger et al. 2018, Chinook salmon size at the time of
their return to native streams were analyzed. This study showed a negative trend in body
size in recent decades. Results from wild fish populations analyzed in Alaska, British
Columbia, Washington, Oregon, and California showed a negative trend and a decline in
17
mean age, with the strongest of this change seen in Alaskan populations. Furthermore, an
interesting find in the relation of size and age showed that the younger individuals (1-2
years) experienced an increase in size where older individuals (4-5) experienced a
decrease in size over time. This result was glaring for individuals in the 4-5 year class for
wild stock and less so for 1-2 year class of wild stock. Specifically this is a 5% decrease
in size at age for 3 year olds, 7% decrease for 4 year olds and, 9% for ages 4-5
(Ohlberger et al., 2018). Most importantly, researchers in this study noted that the decline
in size-at-age was most attributed to size selective harvest. Size selection on Chinook
through commercial fishing is incredibly effective and has shown to produce an
evolutionary response towards smaller average size of fish (Ohlberger et al., 2018). This
is documented through this study and among many others. This study is important as it
shows a growing concern for size selection as well as size-at-age for Pacific Salmon.
While this occurred in Chinook salmon, we may draw similar conclusions for coho
salmon in Big Beef Creek.
Alternative methods
Alternative methods are seen in a study done by Atlas et al. 2021 where it is
argued that by returning to true traditional indigenous methods of fishing such as, reef
nets, dip nets, or fish wheels, fish may be harvested more sustainably by being in control
of selection and having the ability to release wild fish.
Phenotypic plasticity and evolutionary changes:
Reproduction in teleost fish as with other vertebrates is characteristically cyclical.
Where, cyclicity is imposed by the factor that environmental conditions tend to recur
cyclically or seasonally. Furthermore, these reproductive cycles are adaptive for each
18
species and they depend on the evolutionary and ecological niche of the species (Miller,
P. J 1979). Changes in one life state can have extensive repercussions for later life states,
particularly in migratory animals where multiple life-stage transitions are finely tuned to
conditions in radically different environments (Crozier et al., 2008). Furthermore,
plasticity and change in response to climate change are not certain as salmon face a
variety of other stressors from hatcheries, harvest, and dams (Crozier et al., 2008).
Evolutionary changes that occur in salmonids has become increasingly available
as its value towards conservation and management practices becomes more apparent.
This includes changes in migration timing and shaping alternative phenotypes. Data has
shown that over the past century spring and summer Chinook as well as sockeye have
been migrating at earlier times (Crozier et al., 2008). Coho may also deviate from
established run timing in response to environmental conditions such as flow availability
and stream accessibility (Groot & Morgolis 1991). These migration events are timed to
coincide with environmental conditions that maximize individual fitness and many
species including coho salmon will change migration timing to match new conditions
produced by climate change (Kovach et al., 2012). In a study by Kovach et al 2012, pink
salmon in Auke Bay, Alaska were found to have directional selection for earlier
migration. Furthermore, these researchers have discussed there being a strong correlation
between the migration timing and recent climate changes (Kovach et al., 2012). This idea
is further supported by reviewing other studies in the Pacific Northwest specifically
where salmon populations are seen shifting migration timing.
Directional selection in migration timing is also seen in sockeye salmon in Bristol
Bay, Alaska. Quinn et al., 2007 compiled data from returning fish counts and commercial
19
harvest data and observed that the median dates of returning migrating fish became
earlier from 1969 to 2003 in both districts’ studies. This study also notes an interesting
correlation between the harvest timing and the sockeye response where, when harvest
pressure heavily increased at the tail end of the run, the larger the directional selection for
an earlier migration timing was for sockeye (Quinn et al., 2007).
Anderson et al., 2010 also studied directional selection in reproductive timing
specifically for coho salmon in the Cedar River, Washington, while also looking at body
size in relation to reproductive success of the species. These researchers found that
selection on breeding timing changed in form while selection on body size changed in
magnitude (Anderson et al., 2010). Body size as argued in this study and others is among
the most important traits that influence production and survival of offspring (Anderson et
al., 2010). This is because larger individuals often experience more reproductive success.
Specifically, large females can produce more numerous and larger offspring and have
more advantages competition for nesting sites (Anderson et al., 2010). Furthermore, to
loose older and larger individuals of a population result in overall reduction of population
productivity. This is because smaller salmon have lower fecundity, lower offspring
survival and, may not be able to dig deep enough redds to reduce susceptibility to
scouring (Ohlberger et al., 2018).
We know that size selection is a common issue in the harvest of salmon. This is
also documented in a study by Kendall & Quinn, 2017 where 283 years of size selection
on sockeye salmon from 9 different fisheries in Alaska were analyzed. The results
revealed a staggering 72-84% of larger individuals within the population being caught
and leaving smaller individuals to spawn (Kendall & Quinn, 2017). Furthermore, it was
20
seen that a disruptive size selection was more prevalent in males than females and the
length of fishing season also resulted in a greater size selection impact (Kendall & Quinn,
2017). This is due to the fact that fish of different lengths often vary in run timing making
the timing and length of harvest critical factors.
A study using 20 years of data from coho salmon in Oregon showed a
synchronous relationship between adult males and jack males. The proposed explanation
for male fish adopting the jack phenotype depends on growth during early life and in the
freshwater environment with climatic variables such as precipitation providing
supporting evidence (Koseki & Fleming, 2007).
It has also been argued that with the strategy employed by salmon where females
enter the river system first, harvest pressure may be biased towards sex of the fish
(Harvey et al., 2017).
21
Chapter Three Methods:
Streamflow:
Streamflow data from water years 2000-2020 were collected from the USGS
water data archives (USGS Water Resources, 2021) and Kitsap County Public Utility
Department (KPUD) database (KPUD Hydrological Data, 2021). The flow gage was
active under the USGS until 10/4/2012 in which time the KPUD took over. All data
collected had been approved by the respective agencies as opposed to being in the
provisional stage.
Flow gage for Big Beef Creek is located at Latitude 47°38'27”, Longitude
122°47'02" and has a drainage area of 13.8 square miles (Figure 2). The retrieved water
data sets were used to create twenty-year average trends and plotted against daily average
discharge for years 2000-2020 during the months of August to December. This allows for
a twenty-year time period to be analyzed in initial plots for any significant changes
against each year.
22
Figure 2 Map of drainage area to streamflow gauge located on Big Beef Creek.
Weir
A weir is put in place each year from mid-August to the end of December
annually to collect data from the adult coho return migration. Fish are processed within
12 hours of entering the weir (Kinsel & Zimmerman, 2011). Coho are enumerated,
checked for tag status, sex, mark status, length, and scales are taken as an ageing
structure. Only coho with the presence of an adipose fin are passed above the weir, this
allows researchers to identify wild stock vs. hatchery stock fish. Scales and body size
differentiate two different age classes of males with males < 35 cm as jacks and males
between 35 – 45 cm assumed to be jack males (age 3) (Kinsel & Zimmerman, 2011).
23
Harvest
The Regional Mark Processing Center (RMPC) online database was used to
download harvest data based on tag codes from out migrating juvenile coho tagged at the
Big Beef Creek weir. Data was organized by statistical week and from the Marine Area
12, 12B, and 12C fishery only, this is highlighted in Figure 3. This accounts for 92% of
coho salmon harvested in the terminal commercial fishery. Because this study researches
the effects of streamflow on harvest rates, it would not be pertinent to include harvest
data from those caught farther out in other marine areas. Although it is important and
worth mentioning the harvest rates in this study reflect that of a certain marine area and
not of the entire run.
Bar graphs were used to view relationship between harvest and fish that migrated
upstream.
24
Figure 3 Marine 12 fishing area, highlighted in orange cross marking. Map courtesy of
the Washington Department of Fish and Wildlife.
25
Peak adult migration and lengths:
Data collected from the weir at Big Beef Creek containing the total number of fish
that returned to the weir each day were organized by total adults that arrived at the trap.
For this study coho jacks were excluded to maintain consistency among the available
datasets. Furthermore, since I used harvest data to make comparisons and, jack coho are
rarely kept in the commercial fishery, they were excluded. Lengths were then extracted
26
from both the harvest data and total returned adults to the weir each year. Jack and jack
male fish are 100% sampled. These are determined by their size class, either a jack under
35 cm or a jack-male that is between 35-45cm. Once the data has been processed and
scales aged jack males are either found to be jack or males and reassigned. Because of the
differences in sample rates the male sample rate was determined for each year and then
applied to jack males, these were then randomly sampled using R Studio (Version
1.2.5033) (rstudio.com) using the following code:
Df[sample(nrow(df) # of samples needed),]
Juvenile outmigration and tagging
Juvenile outmigration data is collected from April to June annually when smolts
are processed through the juvenile fan trap at Big Beef Creek. Coho smolts are tagged
with Coded Wire Tags (CWT) using wire, Mark IV Tag Injector ©, and V Detector ©
from Northwest Marine Laboratories to ensure placement. Tag numbers were assigned to
groups leaving in early, middle, and late groups which were determined by staff that
processed the fish as the trap based on timing throughout the season. Each group was
assigned different tag codes for migration time. These were not always consistently
divided into three groups due to the number of fish available each year, therefore
resulting in years with only an early and late grouping. For example, 2014 had 63 days of
juvenile tagging and 2013 with the shortest amount of tagging days at 39. The previously
established migration timing categories for juvenile outmigrants began for smolts in
2007, so the available dataset it shorter than the rest of the data available in this study.
This was also used up until the most recent RMPC available data set for harvest. The
average duration for outmigration was 47 days and tags were switched on average 19
27
days for early classified fish, 11 days for middle classified fish, and 21 days for late
classified fish. Some years only early and late groups were established among
outmigrants. The data used to compare return timing came from the terminal net fishery
and was broken up into early, early middle, late middle, and late. This was to delineate
even comparisons in the return data from weeks 38-45, with each category consisting of
two weeks.
Statistical analysis
To calculate a change in time among streamflow from 2000-2020, methods from
Stewart et al., 2005 were followed where:
1. The seasonal fractional flows (SFF) are calculated by the streamflow that
occurs in a given month to the total amount of streamflow for the season each
year. Where a season in this study is defined as September 1 – December 31.
Water year is calculated from October 1- September 29 annually, as defined
by the USGS.
2. The date marking the timing for the center of mass of flow timing (CT) for
each water year was calulated.
The equation for determining the CT for each year is:
CT=∑(qiti)/∑qi
Where qi = daily flow and ti = the number of days from the beginning of the season
(Stewart et al., 2005).
28
A three-year moving average was then calculated for SFF as outlined by Ditmer, 2013 to
enhance statistical robustness of the annual data. The SFF was then compared to that of
the SFF from the years 1971-1979 and 1980-1999, and a two sample t-test was run to
determine significance to historical water data. The water data from the USGS had gaps
from the years 1981-1995, therefore these years were not included in our past comparison
of SFF. The results from CT calculations were then run in a simple linear regression.
Spearmans correlation rank test was used to look at trends in discharge (flow in cfs) and
adult return fish, as neither of these data sets were normally distributed. Spearmans
correlation rank test was also used for weekly harvest and discharge data to determine if
low flows corresponded to a larger harvest rate. Harvest and discharge were also plotted
in simple historgram graphs to visualize this relationship.
Contingency tests were used to determine if migration timing between juveniles and
returning adults was independent or not. For some years where the data did not meet the
assumptions of the contingency test, meaning 20% of cells having less than an expected
frequency of 5 or cells having less than one (Whitlock & Schulter2015), a Fisher’s exact
test was employed. Mosaic plots were also used to visualize comparison of this data.
Average lengths for each year were then calculated and a two-sample t-test was run on
average annual lengths from coho caught in the commercial fishery and those that
returned to the weir to test a potential size selective harvest.
t.test(x~y, data= , alternative = “two.sided”, var.equal = FALSE)
Lastly, a multiple regression was run on center of mass flow timing, seasonal fractional
flow, total harvest, date of peak return, and year variables.
29
lm(Date_of_peak ~ year + var 1 + var 2…, df)
Chapter Four Results:
Initial Daily average flow compared to the twenty-year average showed extreme
variation, with many years having flood events exceeding the twenty-year average.
Graphs are included from 2000, 2010, and 2020 as they represent the beginning, middle,
and end of the dataset (Figure 4). 2003 specifically shows a large spike in discharge at the
end of October. This is likely responsible for the flow bump we see in the twenty-year
average across all plots. For comparison, a flow plot where discharge values that were
within two standard deviations of the average for October in the twenty-year database is
included (Figure 5) to show how this affected the twenty-year average.
30
Figure 4 Streamflow for seasonal duration of August to December for A. 2000, B. 2010,
and C. 2020. Where red is the same in each, being the twenty year average and blue
indicates each years flows
Daily Average Stream Discharge 2000 (cfs) for Big Beef Creek
Discharge (Cubic Feet per Second)
150
A
100
50
0
Aug
Sep
Oct
Nov
Dec
Jan
Date
31
Daily Average Stream Discharge 2010 (cfs) for Big Beef Creek
Discharge (Cubic Feet per Second)
800
B
600
400
200
0
Aug
Sep
Oct
Nov
Dec
Jan
Date
Discharge (Cubic Feet per Second)
Daily Average Stream Discharge 2020 (cfs) for Big Beef Creek
C
400
300
200
100
0
Aug
Sep
Oct
Nov
Date
32
Dec
Jan
Figure 5 Adjusted plot to compare the effects of the 2003 early flooding event
Seasonal Fractional Flow for Big Beef:
The three year moving average for SFF showed a negative trend with up to a 58%
decrease seen Figure 6. The SFF from years 1971-1980 and 1996-2000 showed a slight
decline in flow however much less at 15% decrease, suggesting more significant flow
changes in the last twenty years. The two sample t test between these two data sets also
showed a significant change (p<0.05).
Figure 6 Seasonal fractional flows for A. 2000- 2020 and B. historical data 1971-1980
33
Smoothed SSF (2000-2019)
Seasonal to water year flow fraction
A
. 0.90
y = -0.0058x + 0.5246
R² = 0.0499
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
0
5
10
15
20
Time span in years
Smoothed SSF (1971-1980) (1996-2000)
Seasonal to water year flow fraction
B
. 0.45
y = -0.0015x + 0.3488
R² = 0.0152
0.4
0.35
0.3
0.25
0.2
0
2
4
6
8
10
12
14
16
Time span in years
The centroid of mass flow (CT) was calculated for each year as defined by (Stewart et al.,
2005) and placed in a linear regression as seen below Figure 7.
The CT for the twenty year data set has a weak linear regression with R squared value of
0.006 (Figure 7). However t.test results show significant difference (p<0.05) of mean
flows. The difference from 2000 to 2020 is a delayed CT of 4.32 days. The earliest CT is
seen in 2013 and the latest CT in 2005, the difference between these being 30 days.
34
Center of Mass Flow Timing
Figure 7 Central mass flow timing for Big Beef creek over the years 2000-2020
y=-0.033x+184.9
R2=0.0006
Harvest
On average about 50% of the coho salmon are caught in the commercial terminal
net fishery annually over the past twenty years Table 2 from Marine Area 12.
35
Table 2 Percent of coho harvested in terminal net fishery for each year
Year
% of total run harvested
2001
10%
2002
29%
2003
1%
2004
21%
2005
46%
2006
40%
2007
52%
2008
53%
2009
65%
2010
63%
2011
58%
2012
55%
2013
69%
2014
47%
2015
16%
36
2016
15%
2017
59%
2018
83%
2019
45%
Harvest data is broken into statistical week. Coho are harvested over an 8 week
period from stat week 37 to 45. This is typically the from September to mid-November.
The Spearmans rank correlation used to test the hypothesis on harvest and streamflow
discharge as well as discharge to fish return at trap showed expected relationships Table
3. Harvest to discharge was mostly negatively correlated to stream discharge, with the
exception of years 2001, 2004, 2006, and 2007. The interpretation of this is that harvest
was found to be mostly productive where flows were too low to induce fish migration,
supporting the hypothesis. Alternatively fish return was positively and strongly correlated
with stream discharge with -.08 Spearmans rank in 2006 and -0.9 in 2009 for example.
There was no relationship however found in stream discharge and fish return for years
2001and 2006.
Table 3 Spearmans rank correlation for upstream fish passage and harvest to streamflow
discharge relationships.
37
A negative relationship, indicating correlation supporting the hypothesis was detected for
all years with the exception of four years, 2008, 2011, 2013, and 2017. However, these
are relatively weak positives with 2011 and 2017 having the highest spearmans rank at
0.26 and 0.37 respectively.
Timing
Overall coho salmon recorded as returning to the trap are in decline. The last large return
at Big Beef was seen in 2004 and the simple linear regression for the entire data set in
this study returns R2 of 0.28 Figure 8.
Figure 8 Total adult coho return to Big Beef creek that are passed upstream to spawn
over the past 20 years
38
y=-112.77x+2584.1
R2=0.2888
=
The linear regression of dates of peak migration change for adult coho signals a trend
toward an earlier migration timing. R squared value for the twenty-year peak migration is
0.1557 and the difference from 2000 to 2020 peak is 30 days. The most significant
deviations in from this timing trend change was seen in 2011 and 2013. In 2011 the flow
was below the twenty-year average until close to the end of November. Although, this
overall trend of earlier arriving fish was found to be significant with a p value of 0.04441.
58% of the fish were harvested this year and the spearman’s rank correlation value for
discharge to migration upstream was 0.25. In 2013 the peak return coincided with initial
fish movement upstream at 09/30, a much earlier migration timing which also is
associated with the earliest CT seen in this study. This year 69% of the total run was
harvested and the Spearman’s rank correlation value for discharge to migration upstream
was 0.095. 2013 actually carried a streamflow daily average above the twenty-year
average which would have allowed fish to pass as early as September. These values and
relationships reinforce the hypothesis between flow and fish migration.
39
Figure 9 Peak return of coho at Big Beef Creek that are passed upstream to spawn.
Where date is in julian or calendar day.
Size
Body size in coho salmon found in Big Beef Creek were shown to be decreasing.
Furthermore, the size of individuals caught in the terminal net fishery are larger than
those that arrive at the trap up stream, as well as a more significant negative size trend
seen in male coho (Figure 10). The differences in means of coho salmon are 622 mm in
mean length for those harvested and 581mm in mean length of those that arrive at the trap
(p-value 0.0013)
Figure 10 Box plot of lengths of coho that are caught in the terminal net fishery and
those that are passed upstream. Box plots further browkn down into sex, where harvest
represents those that are caught in the terminal net fishery and female and male are
those that were passed upstream. Lastly, totals represent both male and females
combined that were passed upstream.
40
.
Fish caught in
terminal net fishery
Fish passed
upstream
Simple linear models also show a general decline across both female (Figure 11) and
male (Figure 12) coho, male and female combined (Figure 13) and those that are caught
in the terminal net fishery (Figure 14).
41
Figure 11 Female coho length trend over time, p value of 0.02725.
Female coho passed upstream
y=-2.897x + 623.79
42
r2=0.25555
0.25555
Figure 12 Male coho length trend over time, p value 0.005295.
Male coho passed upstream
r2=0.3753
43
Figure 13 Total length of coho salmon that are passed upstream and arrive to spawning
grounds on Big Beef Creek. P value 0.008715
r2=0.3406
44
Figure 14 Lengths of coho caught in the terminal net fishery, p value 0.04103.
Harvested in terminal net fishery
r2=0.2555
Multiple Regression
Multiple regression results for the variables year, peak migration timing, center of
mass flow, seasonal fractional flow, and total harvest were insignificant. These variables
resulted in weak p values (Table 4) with the exception of length and seasonal fractional
flow. This signifies no direct relationship on each other with the exception of length over
45
time and flow. This possibly shows that over time fish are not only getting smaller but
that larger fish cannot enter the system with little to no flow available.
Table 4: Statistical significance results (p values) of multiple regression.
Peak
Peak
Peak
Peak Migration,
Peak
Migration,
Migration,
Migration,
Year + Length +
Migration,
Year + CT
Year + SFF
Year + Total
SFF
Year + CT,
Harvest
SFF, Total
(y=3.89x+620.16)
Harvest
p value
0.255
0.7207
0.7103
0.04539*
0.526
Juvenile outmigration to adult migration
Coho salmon may follow patterns on early, middle, and late migration. Data from
tagged juvenile coho out-migrants were compared to that caught in the terminal net
fishery in attempt to find a relationship between timing as a strategy at different life
stages. Tag codes allowed for fish to be identified based on their out-migration status and
timing of either early, middle, or late. In some years only early and late classifications
were achieved. A Pearson’s chi-squared test was conducted on all years to determine if
outmigration and return migration were independent of each other or not. Based on the
initial results from the chi-squared test we can assume that these migration timings are
not independent of each other, with the exception of 2008, 2009, and 2016. Results from
test can be seen in Table 8. The three years where the data did not meet the contingency
46
assumptions and a Fishers test was used to determine p-values (2011 p=2.562e-13, 2015
p=0.0059, 2019 p=0.0013).
Chapter Five Discussion
Understanding the relationships and dynamics of combined pressures to coho
salmon at Big Beef Creek is important for future management and recovery goals. The
harvest is often much greater than those fish that arrive at the trap and subsequently
upstream to spawn. In this study I examined 20 years of data and harvest trends with
2018 having the largest number of fish harvested at over 80%. Again, this included only
Marine Area 12 harvest data. Coho salmon from Big Beef are also caught in other mixed
ocean fisheries. Pre-season forecasts are created which allow harvest quotas to be
determined each year. Successful recovery and conservation of Pacific salmon species
while maintaining the availability of unlisted fish for harvest requires a strong
understanding of biological, chemical, physical and hydrological dynamic, each of which
greatly influence population dynamics (B. J. Burke et al., 2013).
Streamflow
The return migration of coho salmon is a phenological trait. When faced with
changes to habitat or climate coho salmon may experience phenological or evolutionary
changes in response. The pressure on salmon by anthropogenic influences such as
habitat loss and increased harvest has been well documented. In this study we attempted
47
to view any resultant changes in migration and timing of adult salmon returning to Big
Beef Creek, WA.
Big Beef Creek discharge showed variability, a slight delay in streamflow, and
increased peak flows. The incredible variability seen in flow among the years included
delays of up to 30 days, which directly affected salmon migration upstream. Flood events
that resulted in increased flow averages often exceeded the twenty-year average. This
resulting data is indicative of a system that is experiencing such fluctuations that are in
support of many hypothesis on climate projections. Global climate changes have been
extensively documented as well as the effect it will have on salmon. Changes in global
climate directly affect the water cycle, precipitation and snowpack (W. D. Burke &
Ficklin, 2017). In addition to these changes, shifts in evapotranspiration (ET) and
infiltration are expected which lead to an altered streamflow quantity and timing (W. D.
Burke & Ficklin, 2017; Stewart et al., 2005). When analyzing trends in streamflow many
researchers have studied the effects of temperature, snowmelt, and thus an earlier spring
flow timing. However, the coastal, low elevated, rain-dominated basins have been less
thoroughly studied (W. D. Burke & Ficklin, 2017). In this study I assessed timing by
calculating the CT as defined by Stewart et al., 2005. This CT measures the timing and
magnitude of streamflow. The difference in CT among the years was found to not
significant therefore future studies may incorporate a larger data set as well as include
precipitation and temperature data as these are also significant environmental factors
relating to salmon. This research may however support projections by Burke & Ficklin,
2017, where a rain dominated Washington water shed was expected to have a higher
percent increase in precipitation and an increase in peak streamflow during the season.
48
This would explain the flow and discharge levels continually exceeding those of the years
prior.
This study specifically analyzed the streamflow of the mainstem Big Beef. We
know that Big Beef has smaller intermittent tributaries which may be favored by coho
salmon as they offer better habitat. Future studies may also analyze the changes seen
specifically in these tributaries as they are experiencing complete droughts that last a
majority of the spawning season.
Pacific Northwest hydroclimatology
The Pacific Northwest interannual variability is driven by the Pacific Decadal
Oscillation and El-Nino Southern Oscillation (ENSO), a phenomenon that includes
interdecadal climate variability (Dittmer, 2013; N. J. Mantua et al., 1997). It has been
well documented that the PDO and ENSO have significant effects on precipitation which
influences streamflow. The PDO at the beginning of our dataset was cold which in
combination with a ENSO can mean above average winter temperatures and below
average winter precipitation (Mantua, 1999). Including PDO and ENSO effects into
changes in streamflow is a confounding reality. We know that it has an effect on
streamflow and future studies may take into account the PDO climate dataset in
comparison to streamflow on Big Beef Creek.
Migration Timing
Migration of coho salmon is variable, where some of the juvenile population
either migrates or differential migration occurs where individuals undertake journeys of
variable distance (Beacham et al., 2019).
49
The timing of migration is critical, and early migration is often associated with
being the best strategy for both adults as well as juveniles. Another frequently employed
strategy for out migrating juveniles is simply in mass, where the number of individuals at
a certain time plays a larger role than time. Ultimately, the actual timing migration is a
response to environmental changes, and the selection on reproductive timing can
accelerate the evolution of other traits (Anderson et al., 2010). If adaptive evolution
occurs population fitness and colonization success may increase (Anderson et al., 2010).
In this study where flow was available, the 2013 adult salmon elected to enter
spawning ground habitat as soon as it was accessible. In contrast to this the 2011 flow
delayed the returning adults two months.
The overall return rates of coho salmon to Big Beef Creek were variable with the
number of adults arriving to the weir ranging from 316 – 4647. While the population has
been able to sustain this drastic drop in peak of fish return to weir is concerning and
certainly shows a population in serious decline. When these adults return upstream there
are multiple strategies that are employed in order to ensure success for offspring. This
includes timing strategies, early arrival and egg deposition at a time that ensures an
incubation period that will last until favorable conditions in the spring for emergence
(McClure et al., 2008). Density of individuals can also influence fitness. When fish
densities are high most highly suitable breeding site locations may be used, limiting and
forcing some females to be less selective about site selection and ultimately choosing less
protected sites (Clark et al., 2014). It is possible that on some years this may have been a
further impact, coupled with harvest pressure and streamflow timing to low numbers seen
in Big Beef Creek. For example, a larger number of fish forced to spawn in the main stem
50
Big Beef as opposed to smaller tributaries. Further studies could be done involving site
selection of adults in comparison to flow and more ideal habitat such as smaller
tributaries with more protective vegetation. It is also possible that low flow results in a
form of habitat loss which has been known to affect evolutionary trajectories (McClure et
al., 2008). We also know temperature greatly affects the health and condition of salmon.
Low stream flows are likely associated with higher temperatures which is a critical factor
when considering the healthy reproduction of salmon, as fitness in warm water is reduced
and high temperatures can even be lethal. Temperatures approach lethal limits regularly
in rivers that are found in Washington and Alaska which affect the times fish can migrate
as well (Crozier et al., 2008; Gale et al., 2011). This thermal stress may even be increased
when fish interact with fishing gear, whether an intentional or non-intentional release.
Migration timing relationships
It was found that the relationship between outmigration timing and returning adult
timing is not independent of each other with the exception of three years. This
relationship is reflected from the harvest data collected on the tag codes placed in
smolt coho from 2007-2019. This is the same data set where marine survival is estimated
for the population annually. On years 2015 and 2016 less than 16% of fish were
harvested. While the contingency test gives reason to support the hypothesis these years
may not be best suited for drawing a relationship since the harvest rates were so low, and
the data is based only on harvest associated with wild tagged coho. The remaining years
in this study harvest rate was 45% and greater, providing better data to represent the
entire run relationship. Furthermore, the varying harvest rate values are expected and
agree with previous estimates which was seen in other literature.
51
The selective pressure on early and middle migrants, in combination with
streamflow delays, means that late returning adults may actually be favored in this
system.
Lastly, since we see so few late category tag groups in our data, this may further
emphasize the relationship and may be indicative that those adults that are passed
upstream are part of this “late” category. This is interesting as the strategy that arriving to
spawning grounds early has been extensively researched and given as an explanation for
a more successful reproduction strategy. Our study suggests that a majority of the
individuals returning still attempt at employing an early migrating strategy, however this
coincides with the most anthropogenic and environmental condition pressures. If late
migrating coho salmon are part of the only population that is migrating upstream without
being harvested at a larger rate, then these are individuals are also now given neutral
advantages to spawning grounds and less competition to individuals who are more fit.
Size
Smaller individuals are left to spawn upstream. This was found in the significant
difference of size among those that were harvested and those that returned to the weir,
likely due to the size selection that occurs as a result of using nets. Small body size in
salmonids is a trait that is often related to low fitness (Beacham et al., 2019). Phenotypic
change that has been recently observed in other studies might largely be due to plastic
(non-genetic) changes (Crozier et al., 2008). The results in this study are also similar to
that in of Kendall & Quinn 2017, where the effects of disruptive size selection were
greater in males than in females.
52
Chapter Six Conclusion:
The results of this study are significant and can contribute to further examination
of population dynamics to wild stock coho salmon found in Big Beef Creek. Understanding
size, distribution, and timing are critical aspects of these dynamics and allow managers to
make better forcasts and predictions regarding the run of coho at Big Beef Creek.
Evolutionary changes may occur in some populations that experience constant stress or
pressure. This research highlighted the multivariate effects on coho salmon and specifically
those that occur in synchronous during the adult lifestage. It is interesting to find an earlier
trend in migration given the pressure the that harvest inflicts on the population, as well as
53
the decrease in seasonal fractional flow found. We know that these play a significant role
in the success of coho spawning, however it may be that the increase in flood eventsthat
occur at the end of the season may pose a larger threat by redd scrouring. These are
certainly areas for further research. Most noteably however, is the decrease in size seen in
fish through this research. Pointing toward significant negative impacts inflicted upon the
species. While I was not able to statistically link all variables in this study together and
build a relationship pattern among them, I was able to show the effects individually and
highlight the many stressors that coho salmon face. As more data becomes available these
effects combined with the historical data may show relationships of the combined stressors.
54
55
Bibliography
Anderson, J. H., Faulds, P. L., Atlas, W. I., Pess, G. R., & Quinn, T. P. (2010). Selection
on breeding date and body size in colonizing coho salmon, Oncorhynchus kisutch.
Molecular Ecology, 19(12), 2562–2573. https://doi.org/10.1111/j.1365294X.2010.04652.x
Atlas, W. I., Ban, N. C., Moore, J. W., Tuohy, A. M., Greening, S., Reid, A. J., Morven,
N., White, E., Housty, W. G., Housty, J. A., Service, C. N., Greba, L., Harrison, S.,
Sharpe, C., Butts, K. I. R., Shepert, W. M., Sweeney-Bergen, E., Macintyre, D.,
Sloat, M. R., & Connors, K. (2021). Indigenous Systems of Management for
Culturally and Ecologically Resilient Pacific Salmon ( Oncorhynchus spp.) Fisheries
. BioScience, 71(2), 186–204. https://doi.org/10.1093/biosci/biaa144
Baker, M. R., Kendall, N. W., Branch, T. A., Schindler, D. E., & Quinn, T. P. (2011).
Selection due to nonretention mortality in gillnet fisheries for salmon. Evolutionary
Applications, 4(3), 429–443. https://doi.org/10.1111/j.1752-4571.2010.00154.x
Bass, A. L., Hinch, S. G., Patterson, D. A., Cooke, S. J., & Farrell, A. P. (2018).
Location-specific consequences of beach seine and gillnet capture on uprivermigrating sockeye salmon migration behavior and fate1. Canadian Journal of
Fisheries and Aquatic Sciences, 75(11), 2011–2023. https://doi.org/10.1139/cjfas2017-0474
Beacham, T. D., Wallace, C., Jonsen, K., McIntosh, B., Candy, J. R., Willis, D., Lynch,
C., & Withler, R. E. (2019). Variation in migration pattern, broodstock origin, and
family productivity of coho salmon hatchery populations in British Columbia,
Canada, derived from parentage-based tagging. Ecology and Evolution, 9(17),
9891–9906. https://doi.org/10.1002/ece3.5530
Beamish, R. J., Sweeting, R. M., Neville, C. M., Lange, K. L., Beacham, T. D., &
Preikshot, D. (2012). Wild chinook salmon survive better than hatchery salmon in a
period of poor production. Environmental Biology of Fishes, 94(1), 135–148.
https://doi.org/10.1007/s10641-011-9783-5
Burke, B. J., Peterson, W. T., Beckman, B. R., Morgan, C., Daly, E. A., & Litz, M.
(2013). Multivariate Models of Adult Pacific Salmon Returns. PLoS ONE, 8(1).
https://doi.org/10.1371/journal.pone.0054134
Burke, W. D., & Ficklin, D. L. (2017). Future projections of streamflow magnitude and
timing differ across coastal watersheds of the western United States. International
Journal of Climatology, 37(13), 4493–4508. https://doi.org/10.1002/joc.5099
Clark, S. M., Dunham, J. B., McEnroe, J. R., & Lightcap, S. W. (2014). Breeding site
selection by coho salmon (Oncorhynchus kisutch) in relation to large wood
additions and factors that influence reproductive success. Canadian Journal of
Fisheries and Aquatic Sciences, 71(10), 1498–1507. https://doi.org/10.1139/cjfas2014-0020
Cochran, S. M., Ricker, S., Anderson, C., Gallagher, S. P., & Ward, D. M. (2019).
56
Comparing abundance-based and tag-based estimates of coho salmon marine
survival. Fisheries Management and Ecology, 26(2), 165–171.
https://doi.org/10.1111/fme.12339
Crozier, L. G., Hendry, A. P., Lawson, P. W., Quinn, T. P., Mantua, N. J., Battin, J.,
Shaw, R. G., & Huey, R. B. (2008). Potential responses to climate change in
organisms with complex life histories: evolution and plasticity in Pacific salmon.
Evolutionary Applications, 1(2), 252–270. https://doi.org/10.1111/j.17524571.2008.00033.x
Davidsen, J. G., Rikardsen, A. H., Thorstad, E. B., Halttunen, E., Mitamura, H., Præbel,
K., Skardhamar, J., & Næsje, T. F. (2013). Homing behaviour of Atlantic Salmon
(Salmo salar) during final phase of marine migration and river entry. Canadian
Journal of Fisheries and Aquatic Sciences, 70(5), 794–802.
https://doi.org/10.1139/cjfas-2012-0352
Dittmer, K. (2013). Changing streamflow on Columbia basin tribal lands-climate change
and salmon. Climatic Change, 120(3), 627–641. https://doi.org/10.1007/s10584013-0745-0
Drenner, S. M., Clark, T. D., Whitney, C. K., Martins, E. G., Cooke, S. J., & Hinch, S. G.
(2012). A synthesis of tagging studies examining the behaviour and survival of
anadromous salmonids in marine environments. PLoS ONE, 7(3), 1–13.
https://doi.org/10.1371/journal.pone.0031311
Ficklin, D. L., Stewart, I. T., & Maurer, E. P. (2013). Climate change impacts on
streamflow and subbasin-scale hydrology in the Upper Colorado River Basin. PloS
One, 8(8). https://doi.org/10.1371/journal.pone.0071297
Fleming, I. a., & Gross, M. R. (2011). Breeding Competition in a Pacific Salmon ( Coho :
Oncorhynchus kisutch ): Measures of Natural and Sexual Selection Author ( s ): Ian
A . Fleming and Mart R . Gross Published by : Society for the Study of Evolution
Stable URL : http://www.jstor.org/stable/. Society, 48(3), 637–657.
Ford, M. J., Hard, J. J., Boelts, B., LaHood, E., & Miller, J. (2008). Estimates of Natural
Selection in a Salmon Population in Captive and Natural Environments.
Conservation Biology, 22(3), 783–794. https://doi.org/10.1111/j
Gale, M. K., Hinch, S. G., Eliason, E. J., Cooke, S. J., & Patterson, D. A. (2011).
Physiological impairment of adult sockeye salmon in fresh water after simulated
capture-and-release across a range of temperatures. Fisheries Research, 112(1–2),
85–95. https://doi.org/10.1016/j.fishres.2011.08.014
Harvey, A. C., Tang, Y., Wennevik, V., Skaala, Ø., & Glover, K. A. (2017). Timing is
everything: Fishing-season placement may represent the most important anglinginduced evolutionary pressure on Atlantic salmon populations. Ecology and
Evolution, 7(18), 7490–7502. https://doi.org/10.1002/ece3.3304
Heard, W. R. (2012). Overview of salmon stock enhancement in southeast Alaska and
57
compatibility with maintenance of hatchery and wild stocks. Environmental Biology
of Fishes, 94(1), 273–283. https://doi.org/10.1007/s10641-011-9855-6
Irvine, J. R. (2009). The successful completion of scientific public policy: lessons learned
while developing Canada’s Wild Salmon Policy. Environmental Science and Policy,
12(2), 140–148. https://doi.org/10.1016/j.envsci.2008.09.007
Kam, J., Knutson, T. R., & Milly, P. C. D. (2018). Climate model assessment of changes
in winter-spring streamflow timing over North America. Journal of Climate, 31(14),
5581–5593. https://doi.org/10.1175/JCLI-D-17-0813.1
Kendall, N. W., & Quinn, T. P. (2017). Quantifying and comparing size selectivity
among Alaskan sockeye salmon fisheries Author ( s ): Neala W . Kendall and
Thomas P . Quinn Published by : Wiley on behalf of the Ecological Society of
America Stable URL : http://www.jstor.org/stable/23213918 RE. 22(3), 804–816.
Kennedy, H. K., James, K. M., & UW Office of Public Archaeology. (1981). Cultural
Resources: Cultural Resource Assesment of the Big Beef Creek Research Facility,
near Seabeck, Kitsap County, Washington. Seattle: Office of Public Archaeology,
Institute for Environmental Studies.
Kinsel, C., & Zimmerman, M. (2011). Intensively Monitored Watersheds: 2009 Fish
Populations Studies in the Hood Canal Stream Complex. August, 94.
http://wdfw.wa.gov/publications/01221/wdfw01221.pdf
Kodama, M., Hard, J. J., & Naish, K. A. (2012). Temporal variation in selection on body
length and date of return in a wild population of coho salmon, Oncorhynchus
kisutch. BMC Evolutionary Biology, 12(1), 1–12. https://doi.org/10.1186/14712148-12-116
Koseki, Y., & Fleming, I. A. (2007). Large-scale frequency dynamics of alternative male
phenotypes in natural populations of coho salmon (Oncorhynchus kisutch): Patterns,
processes, and implications. Canadian Journal of Fisheries and Aquatic Sciences,
64(4), 743–753. https://doi.org/10.1139/F07-046
Kovach, R. P., Gharrett, A. J., & Tallmon, D. A. (2012). Genetic change for earlier
migration timing in a pink salmon population. Proceedings of the Royal Society B:
Biological Sciences, 279(1743), 3870–3878. https://doi.org/10.1098/rspb.2012.1158
Larsen, L. G., & Woelfle-Erskine, C. (2018). Groundwater Is Key to Salmonid
Persistence and Recruitment in Intermittent Mediterranean-Climate Streams. Water
Resources Research, 54(11), 8909–8930. https://doi.org/10.1029/2018WR023324
Law, R. (2000). Fishing, selection, and phenotypic evolution. ICES Journal of Marine
Science, 57(3), 659–668. https://doi.org/10.1006/jmsc.2000.0731
Mantua, N. J., Hare, S. R., Zhange, Y., Wallace, J., & Francis, R. C. (1997). Mantua et al
- PDO PAPER - A Pacific interdecadal climate oscillation with impacts on salmon
production (pp. 1069–1078). JISAO.
58
Mantua, N., Tohver, I., & Hamlet, A. (2010). Climate change impacts on streamflow
extremes and summertime stream temperature and their possible consequences for
freshwater salmon habitat in Washington State. Climatic Change, 102(1–2), 187–
223. https://doi.org/10.1007/s10584-010-9845-2
McClure, M. M., Carlson, S. M., Beechie, T. J., Pess, G. R., Jorgensen, J. C., Sogard, S.
M., Sultan, S. E., Holzer, D. M., Travis, J., Sanderson, B. L., Power, M. E., &
Carmichael, R. W. (2008). ORIGINAL ARTICLE: Evolutionary consequences of
habitat loss for Pacific anadromous salmonids. Evolutionary Applications, 1(2),
300–318. https://doi.org/10.1111/j.1752-4571.2008.00030.x
Miller, P. J. (The Z. S. of L. (1979). Fish phenology: anabolic adaptiveness in teleosts (P.
J. Miller (ed.); 44th ed.).
Nandor, G. F., Longwill, J. R., & Webb, D. L. (2010). Overview of the coded wire tag
program in the greater Pacific region of North America. PNAMP Special
Publication: Tagging, Telemetry and Marking Measures for Monitoring Fish
Populations—A Compendium of New and Recent Science for Use in Informing
Technique and Decision Modalities, 5–46.
http://www.rmpc.org/files/Nandor_et.al.Chap02.pdf%0Ahttp://www.rmpc.org/publi
cations.html
Ogston, L., Gidora, S., Foy, M., & Rosenfeld, J. (2015). Watershed-scale effectiveness of
floodplain habitat restoration for juvenile coho salmon in the chilliwack river,
British Columbia. Canadian Journal of Fisheries and Aquatic Sciences, 72(4), 479–
490. https://doi.org/10.1139/cjfas-2014-0189
Ohlberger, J., Ward, E. J., Schindler, D. E., & Lewis, B. (2018). Demographic changes in
Chinook salmon across the Northeast Pacific Ocean. Fish and Fisheries, 19(3), 533–
546. https://doi.org/10.1111/faf.12272
Quinn, T. P., Hodgson, S., Flynn, L., Hilborn, R., Donald, E., Salmon, S., Nerka, O.,
Quinn, T. P., Hodgson, S., Flynn, L., Hilborn, R., & Rogers, D. E. (2007).
Directional Selection by Fisheries and the Timing of Sockeye Salmon (
Oncorhynchus nerka ) Migrations Published by : Wiley Stable URL :
http://www.jstor.org/stable/40061836 REFERENCES Linked references are
available on JSTOR for this article : You may nee. 17(3), 731–739.
Quinn, T. P., & Phil Peterson, N. (1996). The influence of habitat complexity and fish
size on over-winter survival and growth of individually marked juvenile coho
salmon (Oncorhynchus kisutch) in Big Beef Creek, Washington. Canadian Journal
of Fisheries and Aquatic Sciences, 53(7), 1555–1564. https://doi.org/10.1139/cjfas53-7-1555
Raby, G. D., Hinch, S. G., Patterson, D. A., Hills, J. A., Lisa, A., Raby, G. D., Hinch, S.
G., Patterson, D. A., Hills, J. A., Thompson, L. A., Cooke, S. J., & De, J. (2018).
Mechanisms to explain purse seine bycatch mortality of coho salmon Thompson and
Steven J . Cooke Published by : Wiley on behalf of the Ecological Society of America
Stable URL : https://www.jstor.org/stable/24700327 Mechanisms to explain purse
59
seine bycat. 25(7), 1757–1775.
Russell, J. R., Vulstek, S. C., Joyce, J. E., Kovach, R. P., & Tallmon, D. A. (2018). Longterm changes in length at maturity of Pacific salmon in Auke Creek Alaska. U.S.
Dep. Commer., NOAA Tech. Memo., NMFS-AFSC-384, 28.
Stewart, I. T., Cayan, D. R., & Dettinger, M. D. (2005). Changes toward earlier
streamflow timing across western North America. Journal of Climate, 18(8), 1136–
1155. https://doi.org/10.1175/JCLI3321.1
Vander Haegen, G. E., Ashbrook, C. E., Yi, K. W., & Dixon, J. F. (2004). Survival of
spring chinook salmon captured and released in a selective commercial fishery using
gill nets and tangle nets. Fisheries Research, 68(1–3), 123–133.
https://doi.org/10.1016/j.fishres.2004.02.003
Weitkamp, L., & Neely, K. (2002). Coho salmon (Oncorhynchus kisutch) ocean
migration patterns: Insight from marine coded-wire tag recoveries. Canadian
Journal of Fisheries and Aquatic Sciences, 59(7), 1100–1115.
https://doi.org/10.1139/f02-075
Wigington, P. J., Ebersole, J. L., Colvin, M. E., Leibowitz, S. G., Miller, B., Hansen, B.,
Lavigne, H. R., White, D., Baker, J. P., Church, M. R., Brooks, J. R., Cairns, M. A.,
& Compton, J. E. (2006). Coho Salmon Dependence on Intermittent Steams. Wiley
on Behalf of the Ecological Society of America, 4(10), 513–518.
60
Appendices
61